The labor market has shown clear signs of strain this year, with rising layoff announcements, slower hiring and souring sentiment amongst job seekers. At the same time, a growing list of companies have unveiled ambitious AI plans and, in some cases, explicit AI-related headcount reductions.

This naturally begs the question, are we at the beginning of an AI-driven wave of job losses, or is “AI” simply the new label for conventional belt-tightening?

The emerging evidence points more to the latter. AI is touching the labor market, but so far, its impact is selective and uneven, changing how people work rather than whether they work at all.

1. Adoption is broad, but shallow
   A useful starting point is usage. St. Louis Fed researchers¹ find that 55% of U.S. adults have used generative AI, and over one-third of workers use it for their job. But the intensity of usage is still low. Generative AI is estimated to account for just 2-8% of total U.S. work hours and is primarily used for discrete tasks like drafting emails or code rather than end-to-end workflows.
   
   
2. The “canaries in the coal mine”
   While the aggregate impact is muted, some pockets of the labor market are seeing disruption. A recent study² tracking employment in occupations highly exposed to generative AI finds two patterns:
   1. Since late 2022, early-career workers (ages 22–25) in the most exposed occupations have seen a 13% relative decline in employment compared with peers in less-exposed roles. But mid- and senior-level workers in those fields have seen rising employment.
   2. AI excels at “codified” tasks, the kind of rules-based tasks that can be written in a manual and are often given to juniors, but is much weaker at “tacit” knowledge and soft skills that define more senior roles.
      As a result, pressure is concentrated in routine, digitally native tasks, such as junior software development, customer support and some administrative work.
3. Work must be reliably automatable
   Consider the radiologist—long cited as a prime candidate for AI replacement, yet more in-demand than ever³. The high cost of medical misdiagnosis and regulatory/institutional barriers have meant that hospitals tend to deploy AI alongside radiologists rather than instead of them, and efficiency gains have boosted demand for imaging services overall^4. This is why statistics around AI labor “exposure” can be misleading. Goldman Sachs, for instance, finds that two-thirds of U.S. occupations are exposed to AI, but only 6-7% of workers at risk for full displacement⁵.

AI layoffs in context

We also need to separate recent narrative from scale. U.S. employers have announced over 1.1 million job cuts so far this year⁶. Of that, AI has been cited as a factor in roughly 55,000 job cuts, less than 5% of announced layoffs and equating to ~0.03% of overall employment. Layoffs have risen, but they are still far better explained by government cutbacks, corporate belt-tightening and softening demand than by AI replacement.

AI is fueling a record stock market and an active debate about the future of work, but based on current evidence it is not yet having a material impact on aggregate employment. For investors, the more useful question may be which companies are using AI to supercharge their workforce, rather than simply replace it, in ways that translate into lasting productivity gains.

¹ See Alexander Bick, Adam Blandin, and David Deming (2025), “The State of Generative AI Adoption in 2025”, Federal Reserve Bank of St. Louis.

² See Eric Brynjolfsson, Bharat Chandar and Ruyu Chen (August 2025), “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence”, Stanford Digital Economy Lab.

³ In 2016, Turning Award-winning AI researcher Geoffrey Hinton said “they should stop training radiologists now”, arguing that deep learning would outperform human radiologists within five to ten years. Nearly a decade later, radiology residency programs are offering a record number of positions and the profession was the second-highest-paid medial specialty in 2025.

⁴ See “AI isn’t replacing radiologists” by Works in Progress and Deena Mousa, September 25, 2025.

⁵ See Goldman Sachs, “How will AI affect the global workforce?”, August 13, 2025.

⁶ Source: Challenger, Gray and Christmas. 

7be43aa3-be6e-11f0-8a21-abe89937dd2b